Jython

A TestMaker user posted this to our support email list regarding Jython 
having an upper limit of 64K on any given method. I was wondering if 
this is true? -Frank


From: PitoniakM@...
Date: Sat Nov 1, 2003  12:54:22 PM US/Pacific
To: dev@...
Subject: [Dev] Work around for Maximum Size of Data in a method < 64K
Reply-To: Load and TestMaker Developers List <dev@...>

Folks,

I went to run a rather large script the other day and ran into a 
problem. It apperas that JPython (an i believe java) has a limit of 64k 
of data in any given method, and throws an exception if it is passed. I 
had a rather large hashtable, and ran into this. Anone else seen/dalt 
with this? Wonder if hashtables are less efficient than other storgage 
mechanisms.

many thanks,

mp

Re: [Jython-users] 64 K limit? From: Jim Adrig <jim@tr...> - 2003-11-07 00:46 Attachments: Message as HTML      We have run into this problem also and have found ways of working around it. Of course that is a huge single method for normal coding, but we have an automated system that takes separate scripts (for fields of a printed page) and combines them into a single method for each page. Part of our solution was to shorten the 'standard' method names that are written to the script which wrap the user's scripts; since this worked, I assume the problem is in the size of the source code for each script ? But a user could still go over the limit with careless or extemely complex coding. - Jim On Nov 6, 2003, at 2:20 PM, Frank Cohen wrote: > A TestMaker user posted this to our support email list regarding > Jython having an upper limit of 64K on any given method. I was > wondering if this is true? -Frank > > > From: PitoniakM@... > Date: Sat Nov 1, 2003 12:54:22 PM US/Pacific > To: dev@... > Subject: [Dev] Work around for Maximum Size of Data in a method < 64K > Reply-To: Load and TestMaker Developers List <dev@...> > > Folks, > > I went to run a rather large script the other day and ran into a > problem. It apperas that JPython (an i believe java) has a limit of > 64k of data in any given method, and throws an exception if it is > passed. I had a rather large hashtable, and ran into this. Anone else > seen/dalt with this? Wonder if hashtables are less efficient than > other storgage mechanisms. > > many thanks, > > mp

Re: [Jython-users] 64 K limit? From: Samuele Pedroni <pedronis@bl...> - 2003-11-07 02:49 At 14:20 06.11.2003 -0800, Frank Cohen wrote: >A TestMaker user posted this to our support email list regarding Jython >having an upper limit of 64K on any given method. I was wondering if this >is true? -Frank it's a Java limit, on the code size of methods. Even Sun most recent compiler (1.4.2) does not really work around it: <X.java> public class X { static private java.util.ArrayList list = new java.util.ArrayList(); static public void main(String[] args) { list.add(new Integer(0)); list.add(new Integer(1)); list.add(new Integer(2)); ... list.add(new Integer(5995)); list.add(new Integer(5996)); list.add(new Integer(5997)); list.add(new Integer(5998)); list.add(new Integer(5999)); } } </X.java> D:\exp\java-64k-meth-limit>\usr\jdk142\bin\javac X.java X.java:6: code too large static public void main(String[] args) { ^ 1 error Previously (for example with 1.4.0) compilation would proceed but D:\exp\java-64k-meth-limit>\usr\jdk14\bin\javac X.java D:\exp\java-64k-meth-limit>\usr\jdk14\bin\java X Exception in thread "main" java.lang.ClassFormatError: X (Code of a method longer than 65535 bytes) at java.lang.ClassLoader.defineClass0(Native Method) at java.lang.ClassLoader.defineClass(ClassLoader.java:509) ... which is also what happens with Jython 2.1, for example: <x.py> d = { '0': 0, '1': 1, '2': 2, ... '4995': 4995, '4996': 4996, '4997': 4997, '4998': 4998, '4999': 4999, } </x.py> D:\exp\java-64k-meth-limit>jython x.py Traceback (innermost last): (no code object) at line 0 java.lang.ClassFormatError: org/python/pycode/_pyx0 (Code of a method longer tha n 65535 bytes) at java.lang.ClassLoader.defineClass0(Native Method) at java.lang.ClassLoader.defineClass(ClassLoader.java:509) In such a case not only the module "main" method constructing the dictionary is too large, already the static initializer where all constants (PyStrings,PyIntegers) are constructed is too large. regards.

Re: [Jython-users] 64 K limit? From: Frank Cohen <fcohen@pu...> - 2003-11-14 20:48 Attachments: Message as HTML      Thanks for the reply on this. I was surprised to learn about this - it reminds me of the 64K limit on Intel 8086 memory blocks, but that's just nostalgia speaking. Considering how few people this impacts, I'm not planning to take any action. -Frank On Thursday, November 6, 2003, at 04:45 PM, Jim Adrig wrote: > > We have run into this problem also and have found ways of working > around it. Of course that is a huge single method for normal coding, > but we have an automated system that takes separate scripts (for > fields of a printed page) and combines them into a single method for > each page. > > Part of our solution was to shorten the 'standard' method names that > are written to the script which wrap the user's scripts; since this > worked, I assume the problem is in the size of the source code for > each script ? > > But a user could still go over the limit with careless or extemely > complex coding. > > - Jim > > > On Nov 6, 2003, at 2:20 PM, Frank Cohen wrote: > >> A TestMaker user posted this to our support email list regarding >> Jython having an upper limit of 64K on any given method. I was >> wondering if this is true? -Frank >> >> >> From: PitoniakM@... >> Date: Sat Nov 1, 2003 12:54:22 PM US/Pacific >> To: dev@... >> Subject: [Dev] Work around for Maximum Size of Data in a method < 64K >> Reply-To: Load and TestMaker Developers List >> <dev@...> >> >> Folks, >> >> I went to run a rather large script the other day and ran into a >> problem. It apperas that JPython (an i believe java) has a limit of >> 64k of data in any given method, and throws an exception if it is >> passed. I had a rather large hashtable, and ran into this. Anone else >> seen/dalt with this? Wonder if hashtables are less efficient than >> other storgage mechanisms. >> >> many thanks, >> >> mp >> -- Frank Cohen, PushToTest, http://www.PushToTest.com, phone: 408 374 7426 Free open-source test automation solutions that test and monitor Web-enabled applications, especially Web Services, for scalability and reliability.

Re: [Jython-users] 64 K limit? From: Attila Szegedi <szegedia@fr...> - 2003-11-14 22:19 It's a Java limitation - to speak precisely, the classfile specification (part of JVM spec) says that the code attribute of a single method can have at most 65535 bytes since certain other bits of data that refer to locations inside a method code, as well as the code size field are unsigned 16-bit integers. Attila. ----- Original Message ----- From: Frank Cohen To: Jim Adrig ; jython-users@... Sent: Friday, November 14, 2003 9:48 PM Subject: Re: [Jython-users] 64 K limit? Thanks for the reply on this. I was surprised to learn about this - it reminds me of the 64K limit on Intel 8086 memory blocks, but that's just nostalgia speaking. Considering how few people this impacts, I'm not planning to take any action. -Frank

1 message has been excluded from this view by a project administrator.